trustgraph-cli 2.9.11__tar.gz → 2.10.0__tar.gz

{trustgraph_cli-2.9.11 → trustgraph_cli-2.10.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: trustgraph-cli
-Version: 2.9.11
+Version: 2.10.0
 Summary: TrustGraph provides a means to run a pipeline of flexible AI processing components in a flexible means to achieve a processing pipeline.
 Author-email: "trustgraph.ai" <security@trustgraph.ai>
 Project-URL: Homepage, https://github.com/trustgraph-ai/trustgraph
@@ -8,7 +8,7 @@ Classifier: Programming Language :: Python :: 3
 Classifier: Operating System :: OS Independent
 Requires-Python: >=3.10
 Description-Content-Type: text/markdown
-Requires-Dist: trustgraph-base<2.10,>=2.9
+Requires-Dist: trustgraph-base<2.11,>=2.10
 Requires-Dist: requests
 Requires-Dist: pulsar-client
 Requires-Dist: aiohttp

{trustgraph_cli-2.9.11 → trustgraph_cli-2.10.0}/pyproject.toml

@@ -10,7 +10,7 @@ description = "TrustGraph provides a means to run a pipeline of flexible AI proc
 readme = "README.md"
 requires-python = ">=3.10"
 dependencies = [
-    "trustgraph-base>=2.9,<2.10",
+    "trustgraph-base>=2.10,<2.11",
     "requests",
     "pulsar-client",
     "aiohttp",

{trustgraph_cli-2.9.11 → trustgraph_cli-2.10.0}/trustgraph/cli/invoke_document_embeddings.py

@@ -11,7 +11,10 @@ default_url = os.getenv("TRUSTGRAPH_URL", 'http://localhost:8888/')
 default_token = os.getenv("TRUSTGRAPH_TOKEN", None)
 default_workspace = os.getenv("TRUSTGRAPH_WORKSPACE", "default")
 
-def query(url, flow_id, query_text, collection, limit, token=None, workspace="default"):
+def query(
+    url, flow_id, query_text, collection, limit,
+    attributes=None, token=None, workspace="default",
+):
 
     # Create API client
     api = Api(url=url, token=token, workspace=workspace)
@@ -23,7 +26,8 @@ def query(url, flow_id, query_text, collection, limit, token=None, workspace="de
         result = flow.document_embeddings_query(
             text=query_text,
             collection=collection,
-            limit=limit
+            limit=limit,
+            attributes=attributes,
         )
 
         chunks = result.get("chunks", [])
@@ -83,6 +87,13 @@ def main():
         help='Maximum number of results (default: 10)',
     )
 
+    parser.add_argument(
+        '-a', '--attribute',
+        action='append',
+        default=[],
+        help='Filter attribute as key=value (repeatable)',
+    )
+
     parser.add_argument(
         'query',
         nargs=1,
@@ -91,6 +102,11 @@ def main():
 
     args = parser.parse_args()
 
+    attributes = {}
+    for attr in args.attribute:
+        k, _, v = attr.partition('=')
+        attributes[k] = v
+
     try:
 
         query(
@@ -99,6 +115,7 @@ def main():
             query_text=args.query[0],
             collection=args.collection,
             limit=args.limit,
+            attributes=attributes or None,
             token=args.token,
             workspace=args.workspace,
         )

{trustgraph_cli-2.9.11 → trustgraph_cli-2.10.0}/trustgraph/cli/invoke_graph_embeddings.py

@@ -4,6 +4,7 @@ Returns a list of matching graph entities.
 """
 
 import argparse
+import json
 import os
 from trustgraph.api import Api
 
@@ -11,7 +12,11 @@ default_url = os.getenv("TRUSTGRAPH_URL", 'http://localhost:8888/')
 default_token = os.getenv("TRUSTGRAPH_TOKEN", None)
 default_workspace = os.getenv("TRUSTGRAPH_WORKSPACE", "default")
 
-def query(url, flow_id, query_text, collection, limit, token=None, workspace="default"):
+def query(
+    url, flow_id, query_text, collection, limit,
+    rdf_type=None, attributes=None,
+    token=None, workspace="default",
+):
 
     # Create API client
     api = Api(url=url, token=token, workspace=workspace)
@@ -23,7 +28,9 @@ def query(url, flow_id, query_text, collection, limit, token=None, workspace="de
         result = flow.graph_embeddings_query(
             text=query_text,
             collection=collection,
-            limit=limit
+            limit=limit,
+            rdf_type=rdf_type,
+            attributes=attributes,
         )
 
         entities = result.get("entities", [])
@@ -93,6 +100,19 @@ def main():
         help='Maximum number of results (default: 10)',
     )
 
+    parser.add_argument(
+        '--rdf-type',
+        default=None,
+        help='RDF type IRI to filter results (optional)',
+    )
+
+    parser.add_argument(
+        '-a', '--attribute',
+        action='append',
+        default=[],
+        help='Filter attribute as key=value (repeatable)',
+    )
+
     parser.add_argument(
         'query',
         nargs=1,
@@ -101,6 +121,11 @@ def main():
 
     args = parser.parse_args()
 
+    attributes = {}
+    for attr in args.attribute:
+        k, _, v = attr.partition('=')
+        attributes[k] = v
+
     try:
 
         query(
@@ -109,6 +134,8 @@ def main():
             query_text=args.query[0],
             collection=args.collection,
             limit=args.limit,
+            rdf_type=args.rdf_type,
+            attributes=attributes or None,
             token=args.token,
             workspace=args.workspace,
         )

{trustgraph_cli-2.9.11 → trustgraph_cli-2.10.0}/trustgraph/cli/invoke_row_embeddings.py

@@ -11,7 +11,10 @@ default_url = os.getenv("TRUSTGRAPH_URL", 'http://localhost:8888/')
 default_token = os.getenv("TRUSTGRAPH_TOKEN", None)
 default_workspace = os.getenv("TRUSTGRAPH_WORKSPACE", "default")
 
-def query(url, flow_id, query_text, schema_name, collection, index_name, limit, token=None, workspace="default"):
+def query(
+    url, flow_id, query_text, schema_name, collection, index_name, limit,
+    attributes=None, token=None, workspace="default",
+):
 
     # Create API client
     api = Api(url=url, token=token, workspace=workspace)
@@ -25,6 +28,7 @@ def query(url, flow_id, query_text, schema_name, collection, index_name, limit,
             schema_name=schema_name,
             collection=collection,
             index_name=index_name,
+            attributes=attributes,
             limit=limit
         )
 
@@ -96,6 +100,13 @@ def main():
         help='Maximum number of results (default: 10)',
     )
 
+    parser.add_argument(
+        '-a', '--attribute',
+        action='append',
+        default=[],
+        help='Filter attribute as key=value (repeatable)',
+    )
+
     parser.add_argument(
         'query',
         nargs=1,
@@ -104,6 +115,11 @@ def main():
 
     args = parser.parse_args()
 
+    attributes = {}
+    for attr in args.attribute:
+        k, _, v = attr.partition('=')
+        attributes[k] = v
+
     try:
 
         query(
@@ -114,6 +130,7 @@ def main():
             collection=args.collection,
             index_name=args.index_name,
             limit=args.limit,
+            attributes=attributes or None,
             token=args.token,
             workspace=args.workspace,
         )

{trustgraph_cli-2.9.11 → trustgraph_cli-2.10.0}/trustgraph/cli/nquads.py

@@ -39,6 +39,21 @@ _BAD_IRI = re.compile(r'[\x00-\x20<>"{}|^\x60\\]')
 _LANGTAG = re.compile(r"[a-zA-Z]+(?:-[a-zA-Z0-9]+)*")
 
 
+def valid_language_tag(language):
+    """True when language can be emitted after the '@' of a LANGTAG.
+
+    The wire schema types the language field as a free-form str, so an
+    unusable tag arrives here as data, not as a bug.
+    """
+    return isinstance(language, str) and _LANGTAG.fullmatch(language) is not None
+
+# The body of the LANGTAG production, [a-zA-Z]+ ('-' [a-zA-Z0-9]+)*, which
+# Turtle shares and which is the shape rdflib's Literal(lang=...) accepts.
+# fullmatch, not a trailing $: $ also matches before a trailing newline, so
+# "en\n" would pass and split the quad across two lines.
+_LANGTAG = re.compile(r"[a-zA-Z]+(?:-[a-zA-Z0-9]+)*")
+
+
 def valid_language_tag(language):
     """True when language can be emitted after the '@' of a LANGTAG.
 

trustgraph_cli-2.10.0/trustgraph/cli_version.py

@@ -0,0 +1 @@
+__version__ = "2.10.0"

{trustgraph_cli-2.9.11 → trustgraph_cli-2.10.0}/trustgraph_cli.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: trustgraph-cli
-Version: 2.9.11
+Version: 2.10.0
 Summary: TrustGraph provides a means to run a pipeline of flexible AI processing components in a flexible means to achieve a processing pipeline.
 Author-email: "trustgraph.ai" <security@trustgraph.ai>
 Project-URL: Homepage, https://github.com/trustgraph-ai/trustgraph
@@ -8,7 +8,7 @@ Classifier: Programming Language :: Python :: 3
 Classifier: Operating System :: OS Independent
 Requires-Python: >=3.10
 Description-Content-Type: text/markdown
-Requires-Dist: trustgraph-base<2.10,>=2.9
+Requires-Dist: trustgraph-base<2.11,>=2.10
 Requires-Dist: requests
 Requires-Dist: pulsar-client
 Requires-Dist: aiohttp

{trustgraph_cli-2.9.11 → trustgraph_cli-2.10.0}/trustgraph_cli.egg-info/requires.txt

@@ -1,4 +1,4 @@
-trustgraph-base<2.10,>=2.9
+trustgraph-base<2.11,>=2.10
 requests
 pulsar-client
 aiohttp

trustgraph_cli-2.9.11/trustgraph/cli_version.py

@@ -1 +0,0 @@
-__version__ = "2.9.11"